Book People Archive

Re: Paul Duguid article: "Limits of self-organization: Peer production and quality"

john said:
>    When I occasionally ask about something odd in a DP/PG book,
>    I occasionally get told of places on the Web to find
>    the scans of that book that were used by DP.
>    I don't know if these scans are tracked systematically

d.p. has been saving all its scans all along...

recently, many of those scan-sets have been
collected and placed online, publicly viewable.

it might not be my place to reveal this publicly
-- although i think in a past post i might have
pointed to it -- so i will let d.p. announce that,
if and when they want to do so.

when d.p. does announce the _public_ availability,
i intend to scrape all of their scans and place them
into my own system of "continuous proofreading".

i believe the plan is for d.p. to wait to announce this
until they've gotten _all_ of the scan-sets into place.
i wish they'd release the ones they've already done,
and release the rest as they get them whipped up,
but -- as you might have noticed -- d.p. doesn't
really care much what _i_ wish they would do...      ;+)

also note that some image-providers have "requested"
that d.p. not release their scans, and d.p. often agreed.
(i'll let other people decide whether or not that was wise.)
the upshot is that not all books have scan-sets available.


>    This might not be the ideal solution, in that
>    it doesn't directly show scans and transcriptions side by side.

my "continuous proofreading" system does just that.

i've pointed to the prototype e-books many times, but again:
>   

[Moderator: There was no URL in this submission; I suspect it was
 stripped out by mistake somewhere along the line getting here.
 But I do know of examples of Bowerbird's cyberlibrary editions at 

    http://snowy.arsc.alaska.edu/bowerbird/cyberlibrary.html

 - JMO]

>    But it could give users who get to Gutenberg texts via my site
>    some way of checking text passages that they think might be wrong.
>    So it would hopefully at least be better than what we have now

i'm _not_ of the opinion that imposing even more tasks on
the end-users who are willing to report suspected errors
will be "better than what we have now".   we really need to
_encourage_ people to report _any_ suspected errors to us.

also, with the d.p. system now -- or even a better one --
it might be very difficult to go from a suspected mistake
in a p.g. e-text to the scan that would resolve the matter.

the best you could hope for, i would think, would be
if the .html version had page-numbers that you could
use to start narrowing down the range of the scan...

but you should know that the naming conventions
of the d.p. scan-sets are oftentimes all fouled up --
many do not follow my iron-clad rule that the name
must indicate the page-number that it represents --
so narrowing that range is a trial-and-error affair...

again, this is _not_ a burden we should impose on
(or even suggest to) the rare birds reporting errors.
we want _more_ of them, not less...


>    (and perhaps provide a basis for someone to stage
>    a more integrated presentation.)

i already got it covered john.
(or i _will_ have it covered,
when the scans are freed.)

which is _not_ to say that
you shouldn't proceed on
the system you envision...

your catalog is a huge asset,
and anything that can make it
even better will be even better.

-bowerbird